Visual Perception and Diagnostic Test Truth Tables

Practice vocabulary flashcards covering the stages of visual perception in medical imaging and key metrics derived from diagnostic truth tables.

Computer Aided Diagnosis (CAD)

The use of image processing and machine systems designed to assist or potentially replace clinicians (such as radiologists or nuclear medicine physicians) in diagnosing medical images.

Mental Image Subtraction

A step in the proposed simplistic process of radiologic diagnosis where the clinician compares an image to remembered normal structures to mentally subtract the normal background and isolate abnormal differences.

Three Sequential Processes of Visual Perception

The three successive steps required to visually perceive and assess clinical image data: Detection, Recognition, and Interpretation.

Detection

The necessary first sequential process of visual perception; experiments demonstrate that radiologists miss $20\%$ to $30\%$ of visual clues at this stage.

Recognition

The second sequential process of visual perception; eye-tracking experiments show that visual clues can be physically detected on an image yet dismissed and not recognised.

Interpretation

The final stage of visual perception where a diagnostic decision is made; errors can occur despite clues being detected and recognised due to inexperience, inadequate training, fatigue, prior conditioning, poor judgement, wrong choice, or multiple possible interpretations.

Positive Test Result

A diagnostic test outcome reported as abnormal or undesirable (bad).

Negative Test Result

A diagnostic test outcome reported as normal or desirable (good).

True Positive (TP)

A diagnostic truth table quadrant where the test result is positive and the disease is genuinely present in the patient.

False Positive (FP)

A diagnostic truth table quadrant where the test result is positive, but the disease is absent in the patient.

True Negative (TN)

A diagnostic truth table quadrant where the test result is negative and the disease is genuinely absent in the patient.

False Negative (FN)

A diagnostic truth table quadrant where the test result is negative, but the disease is genuinely present in the patient.

Sensitivity

Also called the true positive fraction ($P(\text{TP})$), the probability of detecting an abnormality when it is present, defined as $\frac{\text{TP}}{\text{TP} + \text{FN}}$.

Specificity

Also called the true negative fraction ($P(\text{TN})$), the probability of a negative (normal) interpretation when no disease exists in the patient, defined as $\frac{\text{TN}}{\text{TN} + \text{FP}}$.

Accuracy

The fraction of all diagnostic interpretations that are correct, incorporating both sensitivity and specificity, defined as $\frac{\text{TP} + \text{TN}}{\text{TP} + \text{FP} + \text{TN} + \text{FN}}$.

Prevalence

The proportion of individuals in a population who have the disease, defined as $\frac{\text{TP} + \text{FN}}{\text{TP} + \text{FN} + \text{TN} + \text{FP}}$.

Positive Predictive Value (PPV)

The probability that a patient has the disease given that they received a positive diagnostic test result, defined as $\frac{\text{TP}}{\text{TP} + \text{FP}}$.

Negative Predictive Value (NPV)

The probability that a patient does not have the disease given that they received a negative diagnostic test result, defined as $\frac{\text{TN}}{\text{TN} + \text{FN}}$.

False Positive Fraction (FPF)

The proportion of disease-free (negative) patients who are incorrectly reported as test-positive, defined as $\frac{\text{FP}}{\text{FP} + \text{TN}}$.

False Negative Fraction (FNF)

The proportion of diseased patients who are incorrectly reported as test-negative, defined as $\frac{\text{FN}}{\text{TP} + \text{FN}}$.